On Mon, 15 Dec 2008, Oliver Neukum wrote:
> Am Montag, 15. Dezember 2008 16:32:59 schrieb Alan Stern:
> > + (sdp->guess_capacity && (sdkp->capacity & 0x01))) {
> > + sd_printk(KERN_INFO, sdkp, "Adjusting the sector count "
> > + "from its reported value: %llu\n",
>
> Maybe this is just me, but if the disk actually has an odd number of sectors
> this code will assume the disk is one sector too big.
The code will make that mistake only if sdp->guess_capacity is set.
Currently the only place the flag gets set is in code you added
yourself, an unusual_devs entry for the USB4500 FW1.04.
> Therefore we should
> still learn about buggy enclosures and add them to the list, so we should
> make a difference between known buggy devices and cases the heuristics
> works, particularly as it may mark a rare cases where a working device
> is truncated. So KERN_INFO seems a bit weak.
No, it should be KERN_INFO. We don't want anything stronger, since
this message will appear every time somebody plugs in a USB storage
device known to have the READ CAPACITY bug. The point of this message
isn't to learn about buggy or non-buggy devices; the point is to let
the user know that the kernel has adjusted the data reported by the
device.
I've got a different scheme for learning about devices where the
heuristic gives the wrong answer. See the "last-sector-hack" patch
just posted to linux-usb.
Alan Stern
--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
